«a paper presented at SPIE’s 1996 Symposium on Smart Structures and Integrated Systems Jennifer L. Pinkerton Anna-Maria R. McGowan Robert W. Moses ...»
Controlled Aeroelastic Response and Airfoil Shaping
Using Adaptive Materials and Integrated Systems
a paper presented at
SPIE’s 1996 Symposium on Smart Structures and Integrated Systems
Jennifer L. Pinkerton
Anna-Maria R. McGowan
Robert W. Moses
Robert C. Scott
Aeroelasticity Branch, NASA Langley Research Center
Hampton, VA 2368 1-0001
February 26-29, 1996
San Diego, CA
Controlled aeroelastic response and airfoil shaping using adaptive materials
and integrated systems Jennifer L. Pinkerton, Anna-Maria R. McGowan, Robert W. Moses, Robert C. Scott, Jennifer Heeg Aeroelasticity Branch, NASA Langley Research Center Hampton, Virginia 23681-0001
Keywords: piezoelectric actuators, active controls, flutter suppression, subcritical response tailoring, buffet load alleviation, neural networks, airfoil shaping
1. INTRODUCTION Since the Wright brothers’ first successful flight, designers have searched for ways to improve both the efficiency and performance of aircraft. Two key research areas that have arisen are controlling aircraft aeroelastic response and controlling airfoil shape. Historically, the use of passive techniques, such as increasing structural stiffness, mass balancing, or modifying geometry, has been the approach for preventing the onset of flutter, buffet, and other undesirable aeroelastic phenomena. The approach originally spawned by airfoil shaping research sought to achieve design-point airfoil camber control through the deflection of conventional wing control surfaces. Although these approaches have been effective, the passive solutions penalize aircraft designs by increasing both weight and cost while decreasing overall performance. However, utilization of existing control surfaces introduces no such additional penalties; thus, secondary applications for these devices have been sought..
When fuel economy became a more important design driver in the early 1970’s, aircraft designers expanded the airfoil shaping-:pplication of the control surfaces to improve the off-design performance of aircraft during the clean-wing phases of flight. One technique called flap scheduling used predetermined flap deflections at specific flight conditions to produce more optimal aerodynamic shapes. I During the past twenty years, considerable research has also been devoted to developing 4-7 active flutter suppression concepts that utilize conventional control surfaces. And more recently, use of these surfaces has been studied for alleviating buffeting. For these latter applications, active use of control surfaces eliminates most of the penalties associated with the passive techniques and provides flexibility since any method of control employed can be varied with configuration or flight condition.
However, there are difficulties with using conventional control surfaces in such secondary applications. These include:
(1) adequately addressing system redundancy, reliability, and maintainability, (2) avoiding compromising the control surface authority available to maneuver an aircraft, (3) obtaining adequate control effectiveness, and (4) not overshadowing ’ performance improvements with increased complexity and structural weight. Because of these concerns, alternatives to utilizing the aerodynamic control surfaces are being studied. The use of adaptive materials as control effectors is one such alternative.
Building upon the rich history of the conventional control surface research efforts discussed above and more recent accomplishments in the application of adaptive materials are four programs in the Aeroelasticity Branch at the NASA Langley Research Center (LaRC) that are attempting to advance the state-of-the-art in controlling both aeroelastic response and airfoil shape. One of these programs is the Piezoelectric Aeroelastic Response Tailoring Investigation (PARTI) program. To date, most applications of adaptive materials to wing flutter suppression have focused on the use of piezoelectric materials. Results available from this application focus primarily on analytical study, with a few reports documenting experimental work.
Weisshaar prclvides a summary of these efforts and a considerable reference list in reference 9. The PARTI program sought to validate and extend the results from such previous wing flutter suppression studies by being the first to wind-tunnel test a relatively large, multi-degree-of-freedom aeroelastic testbed.
The control of‘active flutter suppression systems is the focus of another research effort at NASA LaRC called the Adaptive Neural Control of Aeroelastic Response (ANCAR) program. At this time, numerous studies, like those discussed and referenced in reference 10, have been conducted on the use of neural networks; however, experiments applying this technology to the control of undesirable aeroelastic phenomena are limited. The ANCAR program seeks to experimentally demonstrate, for the first time ever, an adaptive neural-network-based flutter suppression system.
The application of adaptive materials to buffeting alleviation research is relatively recent. One of the first feasibility studies for this application was conducted by Heeg, et al. at NASA LaRC in a table-top wind tunnel. I ’ Another study at NASA LaRC, called the Actively Controlled Response of Buffet Affected Tails (ACROBAT) program, extended this application in 1995, seeking to demonstrate for the first time the effectiveness of using piezoelectric actuators in alleviating vertical tail buffeting at high angles of attack on a large-scale aircraft model.
Airfoil shapin,gstudies incorporating adaptive materials began in the mid-1980’s. The first attempt at active aerodynamic shape control ‘was conducted by Crawley, Warkentin, and Lazarus and involved the use of piezoelectric actuators to generate twist and camber on the surface of a plate. More recent studies, primarily analytical, have focused on assessing the capability of the available adaptive materials to create significant skin deflections. Although promising results have been obtained, man,y researchers have concluded that most adaptive materials lack the strength and out-of-plane displacement capability needed for this application. I 3 - l 5 However, another research effort at NASA LaRC, called the Airfoil THUNDER Testing to Ascertain Characteristics (ATTACH) project, has recently initiated the investigation of a new adaptive material for this application.
The purpose of this paper is to briefly present background information on piezoelectric adaptive materials and to highlight the progress, current status, and future plans of the four programs mentioned above: the Piezoelectric Aeroelastic Response Tailoring Investigation (PARTI), the Adaptive Neural Control of Aeroelastic Response (ANCAR) program, the Actively Controlled Response of Buffet Affected Tails (ACROBAT) program, and the Airfoil THUNDER Testing to Ascertain Characteristics (ATTACH) project.
For the purposies of this paper, the following definitions will be utilized. “Adaptive materials” are materials that alter their shapes when exposed to an external stimulus. When actuators made of these materials are embedded within or affixed to a host structure and then stimulated to create forces on that structure, the result is an “active structure.” An “adaptive structure,” or equivalently a “smart structure,” is then produced when an active structure is commanded by an adaptive control law, which may employ a neural network. l6 An “integrated system” is formed when a control effector (either an adaptive material or a control surface), host structure, sensor, and controller work together to achieve the same functional goal.
2.1 Wind tunr& The programs described herein utilized the two wind tunnels operated within the Aeroelasticity Branch: the Transonic Dynamics Tunnel (TDT), a very large and complex facility, and the Flutter Research and Experiment Device (FRED), a simple table-top device. The TDT, shown in Figure 1, is a 16-foot-by-16-foot test section, closed-circuit, continuous flow wind tunnel capable of testing over a range of stagnation pressures from near zero to atmospheric and Mach numbers from near zero to 1.2, usiing either an air or heavy gas medium, and has the capability to reduce wind speed rapidly in the event of an instability. ” Designed specifically for aeroelastic testing, the TDT has been used for decades to conduct numerous aircraft and rotorcraft aeroelastic and aeroservoelastic tests. The PARTI, ANCAR, and ACROBAT models were tested in the TDT.
The ATTACH model was tested in the FRED wind tunnel, which is shown in Figure 2. FRED is a table-top, open-circuit tunnel with a 6-inch-by-6-inch fully removable acrylic glass test section. Powered by a 2-hp motor, the wind tunnel is capable of operating at a maximum velocity of 38.1 m/s (125 Ws). A single honeycomb screen at the beginning of the contraction duct helps to smooth the flow before it reaches the test section. This tunnel was also used to test earlier ”*’’ aeroelastic applications of adaptive materials.
2.2 Piezoelectric mater&
Piezoelectric materials, which develop a strain when subjected to an electric field and vice versa, are one of the most popular adaptive materials used today. Currently, these materials are divided into two groups, which differ by the direction that they are able to affect a host structure. The first group, commonly called strain actuators, exhibit an in-plane displacement capability. The second group is a new generation of actuators specifically designed to have an out-of-plane displacement capability.
The conventional configuration for an in-plane displacement piezoelectric actuator consists of a single piezoelectric wafer sandwiched between two electrodes. The relationship between an applied electric field and the corresponding behavior of a piezoelectric actuator is well documented. ‘6*18*19 Increased in-plane actuation can be obtained by several means, including grouping multiple wafers into multiple layers. An actuator possessing some out-of-plane displacement capability can also be created by stacking several of the in-plane piezoelectric wafers. 2o Two of the new generation of piezoelectric actuators specifically designed to have an out-of-plane displacement capability are RAINBOW (Reduced And Internally Biased Oxide Wafer) and THUNDER (Thin-Layer Composite-mimorph Piezoelectric Driver and Sensor). Both devices have a monolithic structure and are pre-stressed during fabrication to set the direction of their displacement. RAINBOW, the first of these actuators to be developed, possesses 10 times the displacement capability of the in-plane actuators previously discussed. THUNDER, which was developed in house at NASA LaRC, exhibits an even larger displacement capability.
2.3 Actuator selection methodolow
In general, the selection of an appropriate actuator for use with each program presented in this paper was based on four criteria:
(1) bandwidth, ( 2 ) force, ( 3 ) displacement capability, and (4) ease of application. For flutter suppression and buffeting alleviation, bandwidth, force, and ease of application are the major criteria, and commercially available in-plane actuators suffice. However, for airfoil shaping, the driving criteria are force, displacement capability, and ease of application. Thus, out-of-plane displacement actuators, such as piezoelectric stacks, RAINBOW wafers, and THUNDER wafers, are typically used for this application.
3.1 The Piezoelectric Aeroelastic Response Tailoring Investigation (PARTI) program The Piezoelectric Aeroelastic Response Tailoring Investigation (PARTI) program was the first study in the area of aeroelastic control using adaptive materials to use a relatively large, multi-degree-of-freedom aeroelastic testbed. The PARTI program was a cooperative effort between the NASA Langley Research Center and the Massachusetts Institute of Technology. The objectives of this program were to demonstrate active control of aeroelastic response at subcritical speeds (conditions below the wing flutter speed) and wing flutter suppression using a large-scale aeroelastic wind-tunnel model with distributed piezoelectric actuators and to develop detailed experimental and analytical techniques. In this program, a wind-tunnel model was designed and fabricated, aeroservoelasticanalyses were performed, and the model was ground and wind-tunnel tested in the TDT. As a result of this program, an extensive database of experimental information has been gathered that is instrumental in understanding the many issues associated with applying strain actuation technology to dynamic problems.